sc(Fv)2 is a single chain antibody prepared by linking four variable regions, two light chain variable regions (VL) and two heavy chain variable regions (VH), using linkers and such (Hudson et al., J. Immunol. Methods 1999; 231:177-189).
For example, single chain antibodies having the sequence VH1-linker-VL2-linker-VH3-linker-VL4 or VL2-linker-VH1-linker-VL4-linker-VH3 are known in the art. Depending on the combination of Fvs (molecules in which VH and VL are noncovalently bound), the structure of sc(Fv)2 may exist as one of two types of conformational isomer: a first sc(Fv)2 in which VH1 and VL2, and VH3 and VL4 respectively form an Fv and a second sc(Fv)2 in which VH1 and VL4, and VH3 and VL2 respectively form an Fv.
However, to date, since studies on sc(Fv)2 had been mostly on bispecific sc(Fv)2, there are few if any reports on conformational isomers of sc(Fv)2.
Bispecific sc(Fv)2 is an sc(Fv)2 in which VH1 and VL4, and VH3 and VL2 (or VH1 and VL2, and VH3 and VL4) in the VH1-linker-VL2-linker-VH3-linker-VL4 sequence have variable regions derived from different monoclonal antibodies. In the case of bispecific sc(Fv)2, since VH1 and VL4, or VH3 and VL2 (or VH1 and VL2, or VH3 and VL4) are derived from the same monoclonal antibody, Fv formation is highly efficient and the occurrence of conformational isomers is thought to be to some extent suppressed. In fact, reports to date indicate the utilization of linker length (e.g., 15-5-15 or 15-15-15) in the production of bispecific sc(Fv)2 does not result in a difference in activity (Non-Patent Document 5). Therefore, for bispecific sc(Fv)2, oftentimes, details on conformational isomers are not mentioned. For example, Non-Patent Documents 3, 4, 8, and 9 indicate that correctly combined Fvs exist by confirming the bispecific binding activity; however, they do not provide any quantitative evaluation of the content ratio of the incorrect Fv combinations or the content ratio between the two. Non-Patent Document 6 confirms that monomer-dimer structural conversion resulting from changing the length of the linkers of bispecific sc(Fv)2 (modifying the length of the linkers at both ends or in the middle); however regarding the conformational isomers of sc(Fv)2, the discussion does not go beyond predictions of the molecular structural models. Accordingly, there is no description of the actual content ratio of conformational isomers in the sample or an identification of the structures.
Since conformational isomers of sc(Fv)2 have not been a focus of attention, the regulation of conformational isomers has also not been examined in detail. Non-Patent Document 10 predicts that by making the length of the linkers either 5-15-5 or 15-5-15, a single chain diabody or a bivalent scFv structure will be formed, respectively. This arises from the fact that when the length of a linker in scFv is 12 or less, adjacent VH and VL generally have difficulty in forming an Fv (that is, they difficultly form a monomer). However, it has been reported elsewhere (i.e., in Non-Patent Document 2) that monomers are formed with Fvs in which the linker length is 10 or 5, though in small amount. In fact, even Non-Patent Document 10 acknowledges that the sc(Fv)2 structures obtained using linker lengths of either 5-15-5 or 15-5-15 are not necessarily 100% single chain diabody or bivalent scFv.
As for conformational isomers, since reports made to date only provide predictions of structures arising from Fv combinations and linker lengths, quantitative analyses on the content ratio of conformational isomers and confirmations/demonstrations that the obtained structures are indeed the structures of interest or not have not been carried out. Thus, conformational isomers have not yet been sufficiently evaluated and regulated. Accordingly, in the context of sc(Fv)2 having any linker lengths, it is difficult to predict the content ratio of conformational isomers from Fv combinations and linker lengths. When sc(Fv)2-type molecules are composed of two sets of VH and VL, the existence of two conformational isomers is a problem that needs to be considered.
Regarding low-molecular-weight compounds, many methods for separating optical isomers and geometric isomers are known in the art; however, methods for separating protein isomers have not yet been reported. While numerous methods for separating proteins with one amino acid difference are reported in the literature, to date there have been no reports of methods for separating two conformational isomers possessing a completely identical primary amino acid sequence. Similarly, methods for separating and analyzing conformational isomers of sc(Fv)2, and for confirming the two types of conformational isomer of sc(Fv)2 are not found in the prior art to date.
Since methods for separating conformational isomers of sc(Fv)2 are not yet known in the art, there are accordingly no reports that focus on the differences in activity that arise between the two types of conformational isomers. In bispecific sc(Fv)2, a large difference in activity is easily predicted between correct Fv combinations and incorrect Fv combinations, depending on the conformational isomers. However, in monospecific sc(Fv)2, it is difficult to predict a difference in activity between conformational isomers that are similarly bivalent. In Non-Patent Document 10, the possibility that the activity of the two conformational isomers may differ is not considered; in fact, the activity (binding activity) was measured using a mixture of conformational isomers. This is due to the fact that it is difficult to separate and purify conformational isomers of sc(Fv)2. Accordingly, the conformational isomers could not be prepared in sufficiently high purity to permit a rigorous comparison of the activities.
For an sc(Fv)2 embodiment in which the linker lengths have been modified as well, it has to date been impossible to “identify”, as opposed to making model predictions wherein each of the two conformational isomers are estimated from the linker lengths, or to quantitatively evaluate the content ratio of these conformational isomers. Therefore, quantitative examinations that elucidate the relationship between the sc(Fv)2 linker lengths and the content ratio of conformational isomers have not been carried out to date. Moreover, to date, there are no reports that substantially control the content ratio of conformational isomers by the linker lengths.
Since changing the linker length leads to an alteration in the distance between the two antigen binding sites of sc(Fv)2, the length of the linker(s) may influence the biological activity of the molecule (particularly agonistic activities such as those which dimerize receptors). Therefore, being able to arbitrarily adjust, depending on the type of antigen, the distance between two antigen binding sites by means of varied linker length is desirable. The linker length has been reported to significantly influence the stability (Non-Patent Documents 1 and 2), and in scFv, it is generally known that the shorter the linker, the lower the stability. It is considered to be similar in sc(Fv)2, and it has been reported that, by making the middle linker shorter, dimers are more easily formed (Non-Patent Document 6). To produce highly stable sc(Fv)2, it is desirable for the linker length to be arbitrarily adjustable. Therefore, when developing sc(Fv)2 as pharmaceuticals, being able to isolate the conformational isomer of interest with an arbitrary linker length is considered to be desirable. However, isolation of the two types of conformational isomer, the bivalent scFv and the single chain diabody, for an sc(Fv)2 having arbitrary linker lengths has not yet been reported.
It was previously reported that sc(Fv)2 of anti-human Mpl antibody show TPO-like agonistic activity; it was revealed that sc(Fv)2 has pharmaceutical utility (Non-Patent Document 12). To develop as pharmaceuticals sc(Fv)2 that include conformational isomers, it is necessary to separate and purify the conformational isomer of interest, and to produce a drug substance composed of only one of the conformational isomers; alternatively, when the drug substance is a mixture of conformational isomers, it is necessary to determine the properties of the two types of conformational isomer, and to carry out specification tests to quantitatively analyze the content ratio of each conformational isomer. However, to date, such methods for separating/purifying, quantitatively analyzing, and determining the structure of the conformational isomers of sc(Fv)2 are not known in the art.
Furthermore, while methods for regulating the content ratio of monomers/dimers/trimers/tetramers of scFv using the linker length have been reported in the literature, as noted above, since methods for quantitatively analyzing the conformational isomers of sc(Fv)2 have not been reported in the art, methods for regulating the content ratio of conformational isomers using the linker length are similarly not yet known.    [Non-Patent Document 1] Protein Engineering, 1993, 6(8), 989-995    [Non-Patent Document 2] Protein Engineering, 1994, 7(8), 1027-1033    [Non-Patent Document 3] Journal of Immunology, 1994, 152, 5368-5374    [Non-Patent Document 4] Journal of Immunology, 1995, 154, 4576-4582    [Non-Patent Document 5] PNAS, 1995, 92, 7021-7025    [Non-Patent Document 6] Journal of Molecular Biology, 1999, 293, 41-56    [Non-Patent Document 7] Protein Engineering, 2001, 14(10), 815-823    [Non-Patent Document 8] Journal of Molecular Biology, 2003, 330, 99-111    [Non-Patent Document 9] Protein Eng Des Sel. 2004 April; 17(4):357-66    [Non-Patent Document 10] Clinical Cancer Research, 2004, 10, 1274-1281    [Non-Patent Document 11] Int. J. Cancer, 1998, 77, 763-772    [Non-Patent Document 12] Blood, 2005, 105, 562-566